Abstract 600 Poster Session I, Saturday, 5/1 (poster 195)

Objective: The purpose of this study was to assess the effect of income levels on state-level variations in birth prevalence of Down syndrome. We hypothesized that birth prevalence of Down syndrome may be lower in states with higher levels of per capita income, due to greater access and/or use of prenatal diagnostic services and termination of affected fetuses.

Methods: We used the U.S. 1990 census and the National Center for Health Statistics birth data for the year 1990 to examine maternal age-adjusted birth prevalence of Down syndrome in relation to state-level per capita incomes for African Americans (N=656,987) and Non-Hispanic Whites (N=2,587,922). Statistical analyses included multi-level (hierarchical) logistic regression models, with maternal age and race-specific per capita income measured at the state level used as individual and state-level covariates.

Results: The average per capita income level was $ 8,899 (s.d. $ 1,804) for African Americans and $ 15,106 (s.d. $ 3,887) for Non-Hispanic Whites. There was a statistically significant, maternal age-adjusted variation in state-level birth prevalence of Down syndrome for both African Americans and Non-Hispanic Whites (p-values < 0.001 for significance of state-level random effects). Furthermore, for both racial groups, the maternal age-adjusted risk of Down syndrome was lower as state-level per capita incomes rose. Specifically, a $ 3,000 increase in state-level per capita income was associated with about a 34 % decrease in the risk of Down syndrome for African Americans [age-adjusted odds ratio (OR), 0.66, p < 0.001 for a $ 3,000 increase in per capita income for African Americans] and a 28 % decrease in the risk for Non-Hispanic Whites [age-adjusted odds ratio, 0.72, p<0.001 for a $ 3,000 increase in per capita income for Non-Hispanic Whites).

Conclusions: 1. States with higher levels of per capita income have lower rates of maternal age-adjusted birth prevalence of Down syndrome for both African Americans and Non-Hispanic Whites. 2. Future studies should assess whether this effect of income is due to differences in access and/or use of prenatal diagnostic services in states with different levels of income. In particular, the role of medical insurance, maternal choices and state-level policies regarding prenatal diagnostic and intervention options require further study.